Geospatial Validation of Real Estate Recommendations Using Non-Negative Matrix Factorization

Geospatial Validation of Real Estate Recommendations Using Non-Negative Matrix Factorization | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Geospatial Validation of Real Estate Recommendations Using Non-Negative Matrix Factorization Arash Kolankeh, Ahmed Elsayed This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7797682/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 4 You are reading this latest preprint version Abstract Real estate recommendation systems assist prospective buyers in identifying suitable properties by utilizing historical transaction data. Conventional collaborative filtering techniques often encounter challenges related to data sparsity, which can compromise their performance. Non-Negative Matrix Factorization (NMF) offers a viable method for extracting latent features from sparse user-property interaction matrices. This research implements an NMFbased recommendation model on a Hartford real estate dataset comprising 730 days of transaction records. The recommendations are validated through geospatial analysis, examining their correspondence with established property clusters and ownership records. We evaluate the recommendation hit rate, investigate outlier instances, and explore factors that influence model behavior, including proximity to significant urban features. Our results demonstrate that 87.14% of recommendations aligned with user preferences under a top-5 evaluation metric, while 12.86% were identified as outliers, frequently attributable to shared latent characteristics such as infrastructure proximity or property type. The proposed method effectively handles a matrix sparsity of 98.97%, surpassing the performance of baseline approaches (SVD: 72.3%, PMF: 76.8%, Weighted MF: 79.2%). This case study illustrates the efficacy of integrating NMF with geospatial validation for sparse real estate transaction data, though further validation is necessary to generalize these findings to other markets. Non-Negative Matrix Factorization Real Estate Recommendation Systems Collaborative Filtering Geospatial Validation Matrix Factorization Full Text Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Major Revision 13 Apr, 2026 Reviewers agreed at journal 04 Mar, 2026 Reviewers invited by journal 04 Mar, 2026 First submitted to journal 07 Oct, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7797682","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":600655083,"identity":"e1005708-475a-401c-a718-e5253c908419","order_by":0,"name":"Arash Kolankeh","email":"","orcid":"","institution":"Canadian University Dubai","correspondingAuthor":false,"prefix":"","firstName":"Arash","middleName":"","lastName":"Kolankeh","suffix":""},{"id":600655084,"identity":"8e41afea-ff3a-46f5-af9b-1313fe10acf6","order_by":1,"name":"Ahmed Elsayed","email":"data:image/png;base64,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","orcid":"https://orcid.org/0000-0002-9255-200X","institution":"Canadian University Dubai","correspondingAuthor":true,"prefix":"","firstName":"Ahmed","middleName":"","lastName":"Elsayed","suffix":""}],"badges":[],"createdAt":"2025-10-07 08:56:31","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7797682/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7797682/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":104403885,"identity":"335c691c-9e0b-4952-afc5-a3416b28e346","added_by":"auto","created_at":"2026-03-11 12:19:17","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1306628,"visible":true,"origin":"","legend":"","description":"","filename":"DrArash1.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7797682/v1_covered_6fef3c56-0284-4911-998d-2ae3fed5cd03.pdf"}],"financialInterests":"","formattedTitle":"Geospatial Validation of Real Estate Recommendations Using Non-Negative Matrix Factorization","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":true,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"soft-computing","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"soco","sideBox":"Learn more about [Soft Computing](https://www.springer.com/journal/500)","snPcode":"500","submissionUrl":"https://submission.nature.com/new-submission/500/3","title":"Soft Computing","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Non-Negative Matrix Factorization, Real Estate Recommendation Systems, Collaborative Filtering, Geospatial Validation, Matrix Factorization","lastPublishedDoi":"10.21203/rs.3.rs-7797682/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7797682/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Real estate recommendation systems assist prospective buyers in identifying\nsuitable properties by utilizing historical transaction data. Conventional\ncollaborative filtering techniques often encounter challenges related to data\nsparsity, which can compromise their performance. Non-Negative Matrix Factorization\n(NMF) offers a viable method for extracting latent features from\nsparse user-property interaction matrices. This research implements an NMFbased\nrecommendation model on a Hartford real estate dataset comprising 730\ndays of transaction records. The recommendations are validated through geospatial\nanalysis, examining their correspondence with established property clusters\nand ownership records. We evaluate the recommendation hit rate, investigate\noutlier instances, and explore factors that influence model behavior, including\nproximity to significant urban features. Our results demonstrate that 87.14% of\nrecommendations aligned with user preferences under a top-5 evaluation metric,\nwhile 12.86% were identified as outliers, frequently attributable to shared latent\ncharacteristics such as infrastructure proximity or property type. The proposed\nmethod effectively handles a matrix sparsity of 98.97%, surpassing the performance\nof baseline approaches (SVD: 72.3%, PMF: 76.8%, Weighted MF: 79.2%).\nThis case study illustrates the efficacy of integrating NMF with geospatial validation\nfor sparse real estate transaction data, though further validation is necessary\nto generalize these findings to other markets.","manuscriptTitle":"Geospatial Validation of Real Estate Recommendations Using Non-Negative Matrix Factorization","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-09 05:37:20","doi":"10.21203/rs.3.rs-7797682/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Major Revision","date":"2026-04-14T03:45:19+00:00","index":"","fulltext":""},{"type":"reviewerAgreed","content":"","date":"2026-03-05T00:11:01+00:00","index":0,"fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-03-04T13:28:18+00:00","index":"","fulltext":""},{"type":"submitted","content":"Soft Computing","date":"2025-10-08T02:35:54+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"soft-computing","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"soco","sideBox":"Learn more about [Soft Computing](https://www.springer.com/journal/500)","snPcode":"500","submissionUrl":"https://submission.nature.com/new-submission/500/3","title":"Soft Computing","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"fb06d697-b64b-4e56-8998-287a73d8223a","owner":[],"postedDate":"March 9th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-05-04T10:36:31+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-09 05:37:20","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7797682","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7797682","identity":"rs-7797682","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}